Sensor to pixels: swarm gathering via image-based reinforcement learning

This study highlights the potential of image-based reinforcement learning methods for addressing swarm-related tasks. In multi-agent reinforcement learning, effective policy learning depends on how agents sense, interpret, and process local inputs. Traditional approaches often rely on handcrafted feature extraction or raw vector-based representations, which limit the scalability and efficiency of learned policies concerning input order and size. In this work, we propose an image-based reinforcement learning method for decentralized control of a multi-agent system, where observations are encoded as structured visual inputs that neural networks can process, extracting spatial features and producing decentralized motion control rules. We evaluate our approach on a multi-agent gathering task of agents with limited-range and bearing-only sensing that aim to preserve visibility-graph connectivity during the aggregation. The algorithm’s performance is evaluated against two benchmarks: an analytical solution proposed by Bellaiche and Bruckstein, which ensures gathering success, but progresses slowly, and VariAntNet , a neural network-based framework that gathers much faster, but shows moderate success rates in adverse initial agent configurations. Our method achieves high gathering success, with a speed nearly matching that of VariAntNet . In certain scenarios, it offers a viable alternative to existing approaches.